adalkiran
adalkiran on Hacker News
Hey, thanks! It's a common question that comes from the ML community. The reason is, the main aim of this project is to implement inference from scratch and without using any library. So, Python has lots of abilities to make it easier, and I need some restriction, like using another language/platform, like Go.
Posted by
adalkiran on Sunday August 18, 2024 @05:27PM
from the show-hn dept.
https://github.com/adalkiran/llama-nuts-and-bolts
Hi HN!I’m so excited to show the updated version of my latest open-source project here. The previous version was built for Llama 2 and was now updated to support Llama 3.1 8B-Instruct model.If you are curious like me about how the LLMs (Large Language Models) and transformers work and have delved into conceptual explanations and schematic drawings in the sources but hunger for deeper...
Posted by
adalkiran on Wednesday March 20, 2024 @07:55AM
from the show-hn dept.
https://github.com/adalkiran/llama-nuts-and-bolts
Hi HN!I’m so excited to show my latest open-source project here.I know it’s in a very niche technical domain, but hope you will like my project. Because using Go on Machine Learning and Large Language Models is an interesting experience for me. Please check it out and I’d love to read your thoughts!A holistic way of understanding how LLaMA and its components run in practice, with code and...
Posted by
adalkiran on Sunday February 05, 2023 @07:09AM
from the show-hn dept.
https://github.com/adalkiran/distributed-inference
Hi HN!I’m so excited to show my another open-source project here. It is a PoC project.Distributed Inference is a project to demonstrate an approach to designing cross-language and distributed pipeline in deep learning/machine learning domain, using WebRTC and Redis Streams.This project consists of multiple services, which are written in Go, Python, and TypeScript, running on Docker. It allows...
Posted by
adalkiran on Saturday July 30, 2022 @07:17AM
from the show-hn dept.
https://github.com/adalkiran/inventa-examples
Hi HN!While I’m working on development of a simple and lightweight cross-language distributed deep learning pipeline with WebRTC (which will be open-sourced in the same Github account and announced here in the following weeks), “Inventa” was born as a spin-off library, but it’s designed as a general purpose library to use in any domain.Inventa is a library that supports Go and Python, for...
Hi, approximately it took my two months, including the documentation, without a full-time strict calendar. But before it, I worked on development of a browser-based video conferencing system (using open-source modules, with some other languages except Go), so I have enough know-how about the domain, since the start of the pandemic (of course not deeply as this project).
Thanks for your nice comment :)
You're so right. In most cases, you won't have a problem if you implemented it with existing products, but for 5%, lots of cases may happen :) The complete reason behind WebRTC Nuts and Bolts is this!
Hi, we used for development of a browser-based video conferencing system with my team at the company I used to work for. You're right, WebRTC datachannels may be unreliable. But the nature of UDP is prone to packet loss, unordered packets etc.. To solve these problems, the WebRTC standard offers some error detection/correction/prevention technics. I used it in production for video/audio transfer but I preferred for data streaming (and also signaling) WebSockets.
The WebRTC standard requires some quirks, so it is the reason why my project (WebRTC Nuts and Bolts) was born.
Posted by
adalkiran on Sunday May 29, 2022 @06:27AM
from the show-hn dept.
https://github.com/adalkiran/webrtc-nuts-and-bolts
Hi HN!I’m so excited to show my first open-source project and first post here.I initially started this project to learn Go language, it is an experimental project. The main goal is to track the adventure of a WebRTC stream from start to finish, by debugging the project or tracking the output at console.By trying out this project, you will deep dive into the steps which are taken while starting...